Multi-population Firefly Algorithm

This paper proposes a meta-heuristic Multi-Population Fire-fly Algorithm (MPFA) for single-modal optimization using two multi-population models, i.e., one is based on the island model while the other on the mainland-island model. The unique characteristics of each sub-population is evolved in-dependently and the diversity of the entire population is ef-fectively increased. Sub-populations communicate with each other to exchange information in order to expand the search range of the entire population. In line with this, each sub-population explores a specific part of the search space and contributes its part for exploring the global search space. The main goal of this paper was to analyze the performance between MPFA and the original Firefly Algorithm (FA). Ex-periments were performed on a CEC 2014 benchmark suite consisting of 16 single-objective functions and the obtained results show improvements in most of them